Add documentation about the seed collection service

docs/SUMMARY.md

@@ -3,7 +3,7 @@
 - [QuickPeep]()
 - [Running and Hosting]()
   - [QuickPeep Search]()
-  - [QuickPeep Seed Collection Service]()
+  - [QuickPeep Seed Collection Service](./running/seed_collection_service.md)
   - [QuickPeep Raker]()
   - [QuickPeep Indexer]()
 - [Internals](./internals/index.md)

docs/running/seed_collection_service.md

@@ -0,0 +1,26 @@
+# Running a QuickPeep Seed Collection Service
+
+The QuickPeep web interface has a built-in seed collection page at `/seeds`.
+This is a simple form where anyone can submit a URL and tag it with appropriate tags.
+
+The list of tags can be changed in `quickpeep.ron` (`web` → `seed_collection` → ...).
+
+
+## Retrieving seeds from the Seed Collection Service
+
+First use the `qp-seedcoll-sort` utility included with QuickPeep to sort through the seeds that you have received.
+For each seed that has been received on the web interface, but not yet sorted, you will be given the option to:
+
+- accept (`y`)
+- reject
+    - generic reason (`n`)
+    - because it duplicates an existing entry (`dupe`)
+    - because it's spam (`spam`)
+    - because it's invalid for some reason; e.g. the URL isn't valid (`inv`)
+
+This just marks the seeds in the database but doesn't yet emit them in a format usable by the QuickPeep raker.
+To export the seeds, use `qp-seedcoll-dump path/to/new/file.seed`.
+This command writes a seed file (usable by the Raker) for all the seeds that haven't yet been dumped.
+
+Getting the seeds from your web UI to the Raker is an exercise left for the reader, but I will note that I do this by committing them into a Git repository, with each seed file being dated.
+This also has the benefit of being able to easily publish them for others to use; my seeds are available at https://git.emunest.net/reivilibre/quickpeep_seeds.